A networked host node may generate data to be stored in a primary storage system. The data may be sent to a secondary storage system to back up the data stored in primary storage. Some of the data stored in primary storage may not need to be stored in secondary storage because it may be identical to data previously written to secondary storage. The data may also not be required to be stored in secondary storage because it will not be needed for later recovery of information, or because of other reasons. Storage of the unnecessary data in secondary storage may consume a limited resource (e.g., storage bandwidth, server processing time, storage memory, etc.) which may otherwise be used to store data needed for later recovery. The storage of unnecessary information may therefore result in the loss of information that is needed for later recovery.